Converting bin 1x1 to 2x2

[Neo_uk]

Hi ...

I guess this may be a silly question but is it possible to convert images taken at 1x1 bin to 2x2 bin before stacking ? ..

I've made a rookie mistake where I have shot my luminance data at 2x2 and all my flats and darks for all filters as well ...only problem is all the rgb data I forgot to change the binning to 2x2 so all my calibration frames don't match in dss ...

So is there a way I can batch process all the rgb data from 1x1 to 2x2 to match the rest of the data ?.

Many thanks Chris.

[david_taurus83]

Yes. In Pixinsight. If you don't have it put them in a Dropbox and I'll sort it out for you 👍

[Neo_uk]

hi David..

yes I have pixinsight but not too savvy with it which process would i use ? ...

I've just sorted my luminance data from  last night and with a quick levels and curves in photoshop it looks all weird ...are these reflections from something ?

[vlaiv]

You can do it in ImageJ (Fiji - which is distribution loaded with plugins)

https://imagej.net/Fiji/Downloads

File / import / image sequence ...

Image / transform / bin ...

File / save as / image sequence ...

(in each step you'll get some options, but I think they are self explanatory)

 

[vlaiv]

Just now, Neo_uk said:

yes I have pixinsight but not too savvy with it which process would i use ? ...

Integer resampling

https://pixinsight.com/doc/tools/IntegerResample/IntegerResample.html

Factor of x2 - method average

[Davey-T]

Can you post a bit of equipment info please.

Dave

[Neo_uk]

Sure ..

Altair Wave 102mm triplet 

Planostar 3" .79 reducer 

Zwo oag

Asi1600mm pro

Zwo efw

Badder 1.25" filters 

All shot using asi air pro.

Gain setting 139 unity gain 

60s subs .

I had more or less the same artifacts using my starlight xpress h694 and sx filter wheel ...could it be collimation of the scope or reducer ?.

Many thanks Chris. 

[Davey-T]

Possibly reducer, just try a couple of shots without it to eliminate it.

Dave

[Neo_uk]

Yeah not sure what it is ...I have the correct spacing of 57mm as per the altair data from sensor to the back of the reducer flange ...

So like you have advised I'll try a few shots without it ..

Just a process of elimination I guess.

Many thanks Chris.

[david_taurus83]

16 hours ago, Neo_uk said:

hi David..

yes I have pixinsight but not too savvy with it which process would i use ? ...

 

 

In Pixinsight open the process tab and select image container at the bottom. Add the files you want to convert.

 

 

Not sure if you need to after adding files but I always navigate to the same folder in Base Directory. In Output directory create a new folder somewhere for your new binned files.

 

 

Holding the little triangle at the bottom left drag it into the background to crate a new instance.

 

 

 

You can close the original Image Container now and you will be left with the new instance. If you click on it you will see the original files and folders are still all there. Keep it closed for the next step though.

 

 

As @vlaiv mentions, open Integer Resample.

 

 

Make sure its set to Downsample and 2x2 in this case. Again, hold the little triangle at the bottom and drag the process and drop it over the Image Container instance. PI should now software bin all your RGB data down to 2x2 and populate your new folder.

 

 

See my example. Original files on the left and downsampled on the right. The filename still says 1x1 but the actual file size has been reduced by a factor of almost 4.

 

 

[Neo_uk]

Hi David...

That is so helpful ...and much appreciated!.

Thank you for taking the time to explain and including the screen shots.

Just got to sort out what's causing the weird reflections / artifacts in my images ...looks even worse than the image I posted above when you apply the screen transfer function in pixinsight .

Many thanks again Chris.

[david_taurus83]

Your welcome!

Are those patterns in the flats if you stretch them?

Edited April 5, 2021 by david_taurus83

[Neo_uk]

well i think i might have found my problem ....i looked at the master flat file but couldn't see the patterns but it did seem too bright in the centre ...

so i restacked the luminance channel 1 with the flats id taken and one with out and opened the both in pixinsight and performed the screen transfer function on each file ....ill let you be the judge ! ...

ps... david-taurus83 what you suggested above worked a treat ...  

just got to nail these flat files now me thinks !.

I've included the master flat stretched and unstretched.

many thanks for the help and advise everyone ...much appreciated  ..

Chris.